As data rates increase to meet demand for higher data throughput, signals transmitted through a communication channel are increasing susceptible to corruption by frequency-dependent signal loss of the communication channel, such as intersymbol interference (ISI), and other noise, such as crosstalk, echo, signal dispersion, and distortion. A decision feedback equalizer (DFE) may be used to remove ISI and other noise by using a feedback loop based on previously decided symbols from the received signal. However, a conventional DFE may require an accurate selection of the coefficients of its finite impulse response (FIR) filter. It may be difficult to provide an accurate selection of the coefficients, which may be different for each receive channel. Using a DFE with inaccurate coefficients may reduce the sampling margin (e.g., timing margin and/or voltage margin), and negatively affect the IC performance. Furthermore, a conventional DFE for compensating ISI may consume significant power and require more area.
Accordingly, it would be desirable and useful to provide a way of intersymbol interference compensation addressing some of those concerns.